A longitudinal study of racing Thoroughbreds: performance during the first years of racing



Objectives To develop a profile of the racing careers of Thoroughbred horses in south-eastern Queensland, and to examine factors that affect racing during the first years of racing.

Design and procedure A longitudinal study using the racing records of a defined cohort of Thoroughbred horses that were born on or within 12 months following 1 August 1991. Data collection commenced in 1996 and will continue until all horses cease racing. In this paper two measures of performance were examined: race earnings during the first year of racing and cumulative proportion of horses still racing up to 2 years after their first start. Univariable and multivari-able methods were used to describe and identify factors associated with the performance of these horses.

Results By 31 July 1997, 1804 horses were enrolled in the study including 916, 701, 152 and 35 horses that first raced at 2, 3, 4 and 5 years of age, respectively. During their first year of racing, half the horses earned no more than A 450 from race earnings, and 710 (39.4%) horses earned no money at all. In comparison to poorly performing horses, well-performing horses were more likely to be male, to have started as 2-year-olds and to have had more starts during this year. Of the horses that first started as 2 and 3-year-olds, only 71 and 46% continued racing for at least 1 and 2 years after their first start, respectively. Length of racing life was associated with performance during the first year of racing (number of starts and average earnings per race), and with sex, date of birth and age at first start.

Conclusions The study confirmed a high wastage among racing Thoroughbreds. As expected, premature retirement from racing was linked to poor performance. During the first year of racing, the race earnings of an estimated 1567 (86.9%) horses were insufficient to cover training costs. The 2-year-old racing cohort outperformed the older racing cohorts in each of the performance measures under investigation. Interpretation of this result, and the long-term effects of 2-year-old racing, was limited by the problem of confounding.